TU Darmstadt / ULB / TUbiblio

Empirical Comparison of Data Structures for Line-Of-Sight Computation

Fünfzig, Christoph and Ullrich, Torsten and Fellner, Dieter W. and Bachelder, Edward N. (2007):
Empirical Comparison of Data Structures for Line-Of-Sight Computation.
IEEE Service Center, Piscataway, NJ, In: IEEE International Symposium on Intelligent Signal Processing, [Conference or Workshop Item]

Abstract

Line-of-sight (LOS) computation is important for interrogation of heightfield grids in the context of geo information and many simulation tasks like electromagnetic wave propagation and flight surveillance. Compared to searching the regular grid directly, more advanced data structures like a 2.5 d kd-tree offer better performance. We describe the definition of a 2.5 d kd-tree from the digital elevation model and its use for LOS computation on a point-reconstructed or bilinear-reconstructed terrain surface. For compact storage, we use a wavelet-like storage scheme which saves one half of the storage space without considerably compromising the runtime performance. We give an empirical comparison of both approaches on practical data sets which show the method of choice for CPU computation of LOS.

Item Type: Conference or Workshop Item
Erschienen: 2007
Creators: Fünfzig, Christoph and Ullrich, Torsten and Fellner, Dieter W. and Bachelder, Edward N.
Title: Empirical Comparison of Data Structures for Line-Of-Sight Computation
Language: English
Abstract:

Line-of-sight (LOS) computation is important for interrogation of heightfield grids in the context of geo information and many simulation tasks like electromagnetic wave propagation and flight surveillance. Compared to searching the regular grid directly, more advanced data structures like a 2.5 d kd-tree offer better performance. We describe the definition of a 2.5 d kd-tree from the digital elevation model and its use for LOS computation on a point-reconstructed or bilinear-reconstructed terrain surface. For compact storage, we use a wavelet-like storage scheme which saves one half of the storage space without considerably compromising the runtime performance. We give an empirical comparison of both approaches on practical data sets which show the method of choice for CPU computation of LOS.

Publisher: IEEE Service Center, Piscataway, NJ
Uncontrolled Keywords: Forschungsgruppe Semantic Models, Immersive Systems (SMIS), Geographic information systems (GIS), Data structures, Terrain modeling
Divisions: UNSPECIFIED
20 Department of Computer Science
20 Department of Computer Science > Interactive Graphics Systems
Event Title: IEEE International Symposium on Intelligent Signal Processing
Date Deposited: 16 Apr 2018 09:03
Export:
Suche nach Titel in: TUfind oder in Google

Optionen (nur für Redakteure)

View Item View Item